Reproduction Attempt for "Gender Bias among Professionals"

The is a reproduction attempt for "Gender Bias among Professionals: An Identity-Based Interpretation" by Wu (2020). The effort followed the Guide for Accelerating Computational Reproducibility (ACRE) from the Berkeley Initiative for Transparency in the Social Sciences (BITSS). The reproduction attempt is logged here on the Social Science Reproduction Platform (SSRP).

The attempt uses the author's reproduction package and the paper's appendix. The repo doesn't include the data, but they can be accessed for free at the link.

Description

Installations of R, Python, and Stata will be required to run the analysis and cleaning scripts.

The following is a diagram of all inputs and outputs, where "manual" indicates a manual transcription:

 codebook_dta.log
  └── static_analysis_2019.do
      └── full_sample2019_stata_final.dta
  figure1_nthreads_by_month_first.pdf
  └── figure1_nthreads_by_month_first.xlsx
      └── topic-diff-by-month.csv
          └── static_analysis_2019.do
              └── full_sample2019_stata_final.dta
  figure2_person_by_month_first_1yr.pdf
  └── figure2_topic_diff.R
  figure3a_mean_acad_by_job.pdf
  └── figure3-4_topic_diff_by_job
      ├── summary_stats_0.log
      │   └── static_analysis_2019.do
      │       └── full_sample2019_stata_final.dta
      ├── summary_stats_1.log
      │   └── static_analysis_2019.do
      │       └── full_sample2019_stata_final.dta
      ├── summary_stats_2.log
      │   └── static_analysis_2019.do
      │       └── full_sample2019_stata_final.dta
      ├── summary_stats_3.log
      │   └── static_analysis_2019.do
      │       └── full_sample2019_stata_final.dta
      └── summary_stats_4.log
          └── static_analysis_2019.do
              └── full_sample2019_stata_final.dta
  figure3b_mean_acad_by_job2.pdf
  └── figure3-4_topic_diff_by_job
  figure4a_mean_person_by_job.pdf
  └── figure3-4_topic_diff_by_job
  figure4b_mean_person_by_job2.pdf
  └── figure3-4_topic_diff_by_job
  figure6a_persist_by_length.pdf
  └── figure5-7_dynamic
      ├── AME-by-length.log
      │   └── dynamic_analysis_2019.do
      │       └── full_sample2019_stata_final.dta
      ├── mlogit-full-after-Aug2017.log
      │   └── dynamic_analysis_2019.do
      │       └── full_sample2019_stata_final.dta
      └── mlogit-full-before-Aug2017.log
          └── dynamic_analysis_2019.do
              └── full_sample2019_stata_final.dta
  figure6b_to_person_by_length.pdf
  └── figure5-7_dynamic
  figure7a_persist_by_job.pdf
  └── figure5-7_dynamic
  figure7b_to_person_by_job.pdf
  └── figure5-7_dynamic
  all_cat_2019.csv
  └── prep_for_analysis.py
      ├── EJR0_raw_text_cleaned.csv
      │   └── clean_raw_text.R
      │       └── EJR0_raw_text.csv
      │           └── clean_scraped_data.py
      │               └── scraped_posts.txt
      ├── counts_Nov2017.npz
      │   └── word_count.py
      ├── 0.txt
      │   └── vocab.R
      │       └── cleaned_vocab.csv
      │           └── manual
      │               └── vocab_Nov2017.pkl
      │                   └── word_count.py
      │                       └── EJR0_raw_text_cleaned.csv - CYCLE DETECTED
      ├── 1.5.txt
      │   └── vocab.R
      ├── 1.txt
      │   └── vocab.R
      ├── 2.5.txt
      │   └── vocab.R
      ├── 2.txt
      │   └── vocab.R
      ├── 3.txt
      │   └── vocab.R
      ├── 4.4.txt
      │   └── vocab.R
      ├── 4.5.txt
      │   └── vocab.R
      ├── 4.6.txt
      │   └── vocab.R
      ├── 4.txt
      │   └── vocab.R
      ├── 5.txt
      │   └── vocab.R
      ├── 6.5.txt
      │   └── vocab.R
      ├── 6.txt
      │   └── vocab.R
      ├── 7.txt
      │   └── vocab.R
      ├── 8.txt
      │   └── vocab.R
      ├── exclude_names.txt
      │   └── vocab.R
      └── gender_classifiers.csv
  all_classifiers_2019.csv
  └── prep_for_analysis.py
  author-history-ID-by-full.csv
  └── NBER-post-history-ID (by full).R
      └── nber-author-info.csv
  author-history-ID-by-part.csv
  └── NBER-post-history-ID-round2 (by part).R
      └── nber-author-info.csv
  full_sample2019_stata.dta
  └── manual
      └── full_sample2019_stata.csv
          └── gen_full_sample.R
              ├── full_sample2019.csv
              │   └── merge_sources.R
              │       ├── main_stats_cleaned.csv
              │       │   └── word_count.py
              │       ├── EJR_ALL_gender_classifiers_2019.csv
              │       │   └── prep_for_analysis.py
              │       ├── JMC-history-ID-merged.csv
              │       │   └── manual
              │       │       ├── JMC-history-ID.csv
              │       │       │   └── JMC-post-history-ID.R
              │       │       │       └── jmc_data_gender_nonmissing.csv
              │       │       └── jmc_data_gender_nonmissing.csv
              │       ├── EJR_gender_dataset_Jan2018.csv
              │       ├── author-history-merged.csv
              │       ├── JR0_raw_text_cleaned.csv
              │       └── raw_time_stamp.csv
              ├── EJR_ALL_categories_2019.csv
              │   └── prep_for_analysis.py
              │       ├── gender_classifiers.csv
              ├── full_sample_job_rank.csv
              │   └── job-rank.R
              ├── JMC-history-ID-merged.csv
              │   └── manual
              │       ├── jmc_data_gender_nonmissing.csv
              ├── author-history-merged.csv
              └── complete-nber-author-info.csv

The edits to the original reproduction package are as follows:

• Replace file directory on line 128 of code/dynamic_analysis_2019.do .
• Change file name on line 20 of code/NBER-post-history-ID-round2 (by part).R to NBER/nber-author-info.csv.
• Removed extra file output from code/clean_raw_text.R.